The correct answer is: Deploy two EC2 instances in separate Availability Zones within the same Region, configure database clustering, and implement synchronous replication for automatic failover.

This architecture provides high availability by using two EC2 instances located in different Availability Zones (AZs), reducing the risk of a single point of failure. By clustering the database and enabling replication, the system can automatically fail over to the secondary node if the primary instance becomes unavailable. This approach is commonly used in self-managed database high availability setups on EC2.

Why the other options are incorrect:

Launching a single EC2 instance with AMI backups and CloudFormation introduces manual intervention in the event of failure. While CloudFormation can help re-provision resources, this does not offer automatic failover and downtime is likely during recovery.

Launching EC2 instances across different AWS Regions with replication improves disaster recovery (DR), but not high availability (HA). Cross-Region failover is typically slower and more complex to automate, and does not provide the immediate resilience required for HA.

Using EC2 automatic recovery for a single instance is useful for recovering from hardware issues in the same AZ, but it does not address AZ-wide failures or provide true HA. Additionally, there's still potential downtime during the recovery period.

Why this question matters: This question evaluates your ability to architect resilient EC2-based database solutions. It reinforces the importance of Availability Zones and clustering when RDS is not used and a self-managed database on EC2 is required.

Tip: For EC2-based HA architectures, always look for multi-AZ deployments with clustering or replication. If the solution mentions failover and high availability together, single-instance solutions or DR-only designs are not sufficient.

Full AWS Practitioner Certification Question